KR101202527B1 - Defect observation device and defect observation method - Google Patents

Abstract

Means for storing, in the review SEM, an image set acquired under a plurality of imaging conditions or an image set simulating a plurality of imaging conditions by simulation, means for storing defect position information for each image set, and imaging conditions; Means for storing information on the processing time are provided. Further, means for estimating the predictive defect detection performance and throughput for each of the imaging conditions for the above-described image set from these storage information, and means for displaying the result are provided.

Description

DEFECT OBSERVATION DEVICE AND DEFECT OBSERVATION METHOD}

BACKGROUND OF THE INVENTION Field of the Invention The present invention relates to a defect observation apparatus for observing an image such as a defect generated in a manufacturing process such as a semiconductor wafer, and a technique related to an observation method. In particular, it is easy to set conditions for automatic observation. It relates to technology for.

As the miniaturization of circuit patterns formed on semiconductor wafers proceeds, the effect of defects in the manufacturing process on product yield is increasing, and it is increasingly important to perform process management to prevent defects in the manufacturing stage. Is being done. At present, the yield countermeasure is performed at the manufacturing site of a semiconductor wafer using a wafer inspection apparatus and an observation apparatus. An inspection apparatus irradiates the presence or absence of a defect on a wafer at high speed. The presence or absence of a defect is examined by imaging the state of the wafer surface using optical means (a brightfield type wafer inspection device or a darkfield type wafer inspection device) or an electron beam and processing the image. In the inspection apparatus, since its high speed is important, the amount of image data is reduced by increasing the pixel size of the image to be acquired (ie, lower resolution) as much as possible, and in most cases, defects are detected from the detected low resolution image. Although the presence of can be confirmed, the kind of the defect cannot be determined in detail.

On the other hand, the observation apparatus is an apparatus for acquiring and observing an image with respect to each defect detected by the inspection apparatus in a state where the pixel size is reduced (that is, the resolution is high). In the semiconductor manufacturing process in which miniaturization progresses more and more, the size of the defect to be inspected or observed reaches several tens of nanometer orders, and in order to accurately perform the observation and classification of the defects, the resolution of the nanometer order is increased. This is necessary. Therefore, recently, the observation apparatus (henceforth a review SEM) using a scanning electron microscope is used widely.

In the defect inspection apparatus, the detection sensitivity is recently improved, and a plurality of defects (some hundreds to thousands of defects, sometimes tens of thousands of defects) are detected from one wafer. As a result, the need for the efficiency of the review work for observing the detected defect is higher than ever. In order to cope with such a need, most of the SEM (Scanning Electron Microscope) recently introduced into the market has a function (ADR: Automatic Defect Review) or an image which automatically captures an image of a defect position detected by an inspection apparatus. Automatic Defect Classification (ADC) is included.

For example, Patent Document 1 discloses a related art regarding the function of automatic image collection and automatic defect classification in a review SEM. This patent document 1 describes the structure of a review SEM, the function and operation sequence of automatic image collection and automatic defect classification, the display method of the acquired image, the classification result, etc.

Patent Document 1: Japanese Patent Application Laid-Open No. 2001-331784

First, the outline | summary is demonstrated about the automatic collection function of the defect image in the above-mentioned review SEM, and the subject which this invention solves is demonstrated next.

The defect image automatic collection function in the review SEM is a function of automatically acquiring images of the respective portions by inputting the defect position information obtained as a result of the defect inspection in the inspection apparatus for the wafer to be observed as described above. The basic sequence of this function is

(1) Movement of the sample mounting stage for the defect to be observed to enter the imaging visual field of the review SEM by using the position information of the defect detected by the inspection apparatus

(2) It is high magnification (about 50,000-200,000 times) image pick-up of the defect site by review SEM, and a function is realized by repeating these processes to each defect.

However, in image pickup of a defect site, it is necessary to consider the coordinate error of the inspection apparatus. This is because a normal inspection apparatus may include coordinate errors of several micrometers to several ten micrometers. If a high magnification (for example, 50,000 to 200,000 times) image is taken directly at a defect coordinate position, it is possible to obtain an image within the field of view. This is because there is a possibility that a defect does not enter. Therefore, in this case, an image is picked up in a wide field of view (for example, 15 micrometers-10 micrometers) in the area | region where a defect exists in a review SEM initially, and the defect position detection is performed with respect to the image by image processing. It responds by imaging an image in a narrow field of view (for example, 10 micrometers-4 micrometers) so that a detected position may become the center of a visual field. In this case, therefore, imaging of the image by the review SEM is performed twice in the above-described basic sequence.

As one method for realizing the defect position detection in the above-described sequence, there is a method of comparing an image (defect image) in which no defect exists and an image (defect image) photographing a defect portion. In the semiconductor wafer, since the same circuit pattern is repeatedly formed in units of chips, when detecting a defect position by this comparison method, an image of an adjacent chip portion at the same position as the defect present portion is usually used as a reference image. In this case, therefore, in addition to the above-described sequence, image pickup processing of the reference image and stage movement therefor are further required. Further, at the time of image acquisition, image quality adjustment processing such as focusing and brightness adjustment processing is also required.

As described above, the image quality adjustment processing is performed in each case of performing a plurality of image pickups for one defect, such as a sequence for acquiring a reference image other than the defect image or a sequence for imaging the defect image in two different visual fields. May be required.

By the way, throughput is an important performance index for this defect image automatic collection function. It is expected that the higher the throughput, the more defects can be observed in unit time, and the higher the accuracy of the grasp of the defect occurrence situation and the determination of the countermeasure method. In order to improve the throughput, in addition to the reduction of the time required for the reduction of the stage movement and the quality control functions such as focusing, the reduction of the image pickup time alone is required.

There are various approaches to shortening image pickup. For example, reduction of the number of images used for an image averaging process is mentioned. Since the SEM image has a lot of shot noise and a bad S / N ratio, it is common to acquire an image having a high S / N ratio by imaging the same location a plurality of times and averaging those images. If the image used for this averaging process is reduced, processing time will be shortened. In addition, an increase in the amount of current (hereinafter, probe current) of the electron beam irradiated to the sample is also effective for shortening the imaging time. It is because an image with a higher S / N can be obtained even if the amount of probe current is large, even if the same number of averages is obtained. In addition, when the image size (number of pixels) is reduced, the effect of shortening the collective time of the image capturing time by reducing the number of pixels can be expected, as well as shortening the image processing time and transferring the image in the system. • The effect of shortening the time for storage can also be expected, and as a result, it is effective in improving throughput.

On the other hand, the reduction in the number of images for the averaging process described above, the increase in the probe current, and the reduction in the image size become more difficult to detect from the viewpoint of the defect detection processing for the image acquired at a low magnification (i.e., wide field of view). Act in the direction. For example, the reduction of the number of images causes defect detection from a lower S / N image, so that there is a high possibility of false detection using noise or the like as a defect.

In addition, the increase of the probe current may cause charging phenomenon occurring on the surface of the sample or attachment of the contamination to the sample by electron beam irradiation. In this case, there is a difference in the brightness of the image even at a non-defective part. As a result, there is a high possibility of misdetecting the top part as a defective part. In addition, in the case where the field of view of the imaging area is fixed, the reduction of the image size is equivalent to the enlargement of the pixel size, so that it is difficult to automatically detect a defect of a size close to or less than the pixel size, thereby reducing the minute defect. There is a risk of missing things.

As described above, the change of the image pickup condition for the purpose of throughput improvement may increase the risk of lowering the defect detection performance. Since the relationship between the throughput and the defect detection performance is also different in the process and the type of the wafer to be inspected, in reality, the task of finding the imaging conditions satisfying both the throughput and the defect detection performance is dependent on the manual operation of the operator. . As a result, when conditions need to be set frequently, such as a small quantity production line of many kinds, there existed a problem that the operation | work requires huge time.

An object of the present invention is to solve the above problems and to provide a defect observation method and apparatus for satisfying high detection performance while maintaining high throughput.

In order to achieve the above object, in the present invention, the review SEM includes means for storing image sets acquired under a plurality of imaging conditions, means for giving and storing defect position information for each image set, and imaging conditions. For each parameter to be configured (optical parameters such as acceleration voltage and probe current, and imaging parameters such as image size, number of averages, etc.), the corresponding relationship between the setting candidate value and processing time when each candidate value is set is stored. The means were installed.

Further, means for setting a plurality of imaging conditions by a combination of candidate values for each parameter constituting the imaging conditions, and means for calculating defect detection performance and throughput performance for the plurality of imaging conditions. Further, means for automatically selecting one or a plurality of imaging conditions from a plurality of imaging conditions was provided on the basis of defect detection performance and throughput values. Moreover, the function which displays the defect detection performance and throughput value which were calculated for each of several set imaging conditions in correspondence, and the means which selectively display the imaging conditions automatically selected from among several set imaging conditions were provided.

According to the present invention, the user can easily grasp the relationship between the content of the condition setting and the performance indicator with respect to the two performance indicators of the defect detection performance and the throughput performance that vary depending on various conditions of the automatic image collection function set in the apparatus. Will be. As a result, the condition setting at the time of automatic review can be easily performed.

1 is a block diagram showing the configuration of a defect observing apparatus according to a first embodiment of the present invention.
2 is a block diagram showing the configuration of an image pickup unit of a defect observation system.
3 is a flow chart for explaining the flow of processing of defect image collection.
Fig. 4 is a flowchart for explaining the flow of the imaging condition evaluation method according to the first embodiment of the present invention.
5 is a diagram illustrating an example of a list of imaging condition parameters.
6 is a front view of a teaching screen of defect information according to the first embodiment of the present invention.
7 is a diagram showing an example of a correspondence relationship between a candidate value of an imaging condition parameter and a processing time;
8 is a front view of a screen on which an imaging condition evaluation result is displayed;
9 is a front view of a screen on which an imaging condition evaluation result is displayed.
10 is a block diagram showing a configuration of a defect observing apparatus according to a second embodiment of the present invention.
11 is a flowchart for explaining the flow of the imaging condition evaluation method according to the second embodiment of the present invention.
12 is a diagram illustrating an example of a list of imaging condition parameters.
Fig. 13 is a front view of a screen example for displaying an imaging condition evaluation result according to the third embodiment of the present invention.

Hereinafter, the specific implementation form of the defect observation method which concerns on this invention is demonstrated.

≪ Example 1 >

1 shows a configuration diagram of a defect observing apparatus according to the present invention. The apparatus includes an image capturing unit 101 for capturing a defective image, an overall control unit 102 for controlling the entire apparatus, and an input / output unit 103 having display functions such as input of various commands to the apparatus and processing results. , The recipe storage unit 104 for storing various setting conditions (recipe) when executing the automatic collection function, the image storage unit 105 for evaluation for storing the image for evaluation used for setting the recipe, and the recipe for setting A recipe evaluation section 106 for evaluating the defect detection rate and throughput, and a processing time data storage section 107 storing processing time data, which is data necessary for performing the calculation of the throughput varying according to the setting conditions. The image pickup unit 101 has a function for acquiring an image of a localized portion of the sample. In addition, in the present embodiment described below, the case where a scanning electron microscope (SEM) is used as this image pickup unit will be described. However, the present invention is not limited to this embodiment, and optical defect images are acquired. A means may be sufficient.

2 is a diagram showing an example of the configuration of the image pickup unit 101 using SEM. The sample wafer 201 is mounted on the movable stage 202. The electron beam 215 emitted from the electron source 203 and accelerated by the extraction electrode 204 is focused by the capacitor lens 205 or the aperture 206 objective lens 207 and is incident on the sample surface. Secondary electrons, reflected electrons, and the like generated from the surface of the sample are detected by the detector 208 and photoelectrically converted, and the amount thereof is converted into digital data by an analog-to-digital (A / D) converter 209 or the like. By the two-dimensional scanning of the electron beam 215 on the sample by the deflector 210, a two-dimensional digital image on the sample can be obtained, and the obtained image is stored in the image storage section 220. FIG. Each part is connected to all the control parts 102 via the bus 116.

In this apparatus, changing the area (field of view) scanned by the deflector 210 means changing the field of view (= magnification) at the time of image pickup. In addition, the magnitude of the conversion clock interval (sampling interval) when converting into a digital signal in the analog-to-digital (A / D) converter 209 corresponds to the magnitude of the image size of the digital image to be acquired. For example, finer sampling intervals in the same field of view are equivalent to smaller pixel sizes, and finer defects can be captured as images than in this case. Various setting conditions in such image acquisition and other optical conditions (image acceleration voltage, current amount (probe current amount), etc. of the electron beam 204 to be irradiated), etc., in the image capturing are instructed from the overall control unit 102. Is set via the bus 116.

Next, the processing steps of the automatic defect image collection function executed in the apparatus shown in FIG. 1 and the parameter contents (recipe contents) of the condition setting to be set will be described. In addition, two performance indices (defect detection performance and throughput) to be considered when setting conditions are also described.

Here, the defect image collection function is a function which automatically collects an image such as a defect existing on a sample or a location where a defect is suspected to occur, by the image pickup unit 101. The coordinate position of the defect to be imaged is input from the outside. Specifically, the defect inspection apparatus for the purpose of acquiring the presence position of the defect on a sample and the shape of the circuit pattern formed on a sample are estimated, and the location which may form a pattern different from a desired pattern is identified. From an exposure simulator or the like.

3 shows a step of image collection. This figure is an example of an image collection sequence, and it is anticipated that the wide field image containing a defect site | part and the same circuit pattern as a defect site | part are formed using the defect coordinate obtained by the defect inspection apparatus about any defect. Steps for acquiring a total of three images of the wide field of view image of the reference site to be used and the image of the narrow field of view of the defective portion are shown.

In the flow shown in FIG. 3, first, the stage is moved so that the reference site enters the field of view of the image pickup unit (T1). Next, an image of a wide field of view (for example, a field size of several tens of micrometers in both the longitudinal direction and the transverse direction) is acquired (T2). When a defect is detected by die comparison, a reference site is a location shifted by one chip with respect to the defect coordinates. In addition, this processing includes image quality adjustment processing such as focusing (auto focus) and image brightness value adjustment for acquiring a clear and high quality image. Thereafter, the stage is moved so that the defective portion enters the field of view (T3). Next, a wide-field defect image is imaged (T4). Thereafter, the defect positions are detected by comparing the two obtained wide-field images (T5). Thereafter, a defective image of a narrow field of view (e.g., several micrometers in both the longitudinal direction and the transverse direction) is obtained for the detected position (T6). The above is the imaging sequence for one defect, and this process is performed sequentially for the plurality of defects on the wafer.

In addition, the purpose of acquiring the reference image is to perform defect position detection from the defect image of the wide field of view by comparison with the reference image as described above. Although the same circuit pattern is repeatedly formed in the circuit pattern of a semiconductor like the memory cell part of a flash memory device, etc., for such a repeating pattern, as mentioned above, the reference image is divided into one chip with respect to a defect coordinate. It is possible to synthesize | combine by using the repeatability of a circuit pattern from the defect image obtained by image | photographing the part containing a defect, rather than the method of imaging and acquiring the part without a defect of the circuit pattern of the shifted semiconductor. Thus, the position of a defect can be detected by comparing the reference image created by combining the defect image of a wide field of view from this defect image. Therefore, if any method can determine in advance that the location where the defect has occurred is a repetitive pattern portion such as a memory cell portion, the acquisition of a reference image or the stage movement resulting therefrom becomes unnecessary. The reason for acquiring images of two kinds of fields of view of the wide field and narrow field of view is that, as described above, only the image of the narrow field of view is acquired due to an error in defect coordinates output from the inspection apparatus, a movement error of the stage, or the like. This is because the inclusion of a defect in the field of view may not be guaranteed.

From the processing sequence of the automatic image collection described above and the image pickup principle of the review SEM described above, the following five items are included in the processing parameters to be set in order to realize the defective image automatic collection function.

(1) acceleration voltage

(2) probe current

(3) The number of images (each of wide field image and narrow field image) to use for averaging processing

(4) Image size (each of wide field of view image and narrow field of view image)

(5) Viewing size (each of wide field of view image and narrow field of view image)

These condition values need to be stored in the recipe storage unit 104 as a recipe prior to the execution of the automatic image collection.

By the way, in the automatic collection function of a defect image, defect detection performance and throughput are important performance indexes. First of all, it is a defect detection performance, but this means the accuracy of a process of detecting a defect position from a defect image of a wide field of view. As a result of the failure of the defect detection, if a portion other than the defective portion is incorrectly detected as a defect, the image of the narrow field of view photographing the portion is of course meaningless. Therefore, the defect detection rate is usually required to have an accuracy of 95% or more.

Another important performance indicator in the automatic collection of defect images is throughput, which is the number of image collection defects per unit time. In order to improve the throughput, the processing time in each step shown in Fig. 3, specifically, the narrow field of view, the image capturing time of the wide field of view, the movement time of the stage, the time of the defect detection processing, the time of the image quality adjustment processing, etc. You need to cut it.

Next, the recipe setting method in this invention is demonstrated. 4 shows a processing flow. First, image data acquired by a plurality of imaging condition sets in which different values are set in the above-described processing parameters (1) to (5) is acquired, and stored in the evaluation image storage 105 (S1).

As one of these specific implementation forms, the evaluation image storage unit 105 holds list data of setting candidate values of each parameter as shown in FIG. 5 in a table format. The overall control unit 102 creates a plurality of imaging condition sets by the combination of candidate values of each parameter, and images the image by the image imaging unit 102 with the contents of each imaging condition set.

In the example shown in FIG. 5, as an example of a setting candidate value for a parameter, (1) three acceleration voltages, (2) three probe currents, (3) four averaging processes, and (4) four pixel sizes (5) The case of four types of visual fields is shown. Since the number of candidate values of the parameters varies somewhat depending on the contents set in the table, when the candidate values for each parameter are increased, the total number of the combinations explodes. Therefore, in order to shorten the acquisition time of an evaluation image, it is effective to narrow down the kind of parameter in advance, or to reduce the number of candidate values. On the stage 202, a sample wafer for collecting image data is mounted in advance, and the entire control unit 102 is based on the imaging conditions generated by the combination of candidate values for the above-described parameters. In the illustrated processing step, the image capturing unit 101 instructs to acquire an image. Although the defect which image | photographs is practical of several to several dozen, it is not limited to this. The picked-up image is stored in the image storage unit 210.

Next, the position of each defect part in an image is taught about the defect image of the wide field of view of the acquired image data for evaluation (S2). 6 is an example of a display screen of the input / output unit 103 that performs the teaching process. The thumbnail portion 601 is a portion where thumbnail images collected are displayed. From the image storage section 210, the image data picked up under the same conditions in the image picked up the sample wafer is read out, and a list thereof is displayed here. In the example of FIG. 6, the code | symbol 6011-6014 are shown as an example of the collected defect image. By selecting an arbitrary image from the reference numerals 6011 to 6014 in this area by the mouse cursor 605, the image can be enlarged and displayed in the teaching area 602.

In this teaching area 602, the position of the defect site | part 603 is registered with the mouse cursor 605 about the image displayed on the screen. Specifically, the defect definition area 604 which means the center position of a defect (+ in drawing) and the range of the defect (0 in drawing) is defined and registered by operating the mouse cursor 605 on the screen. Teaching is performed on the acquired image data by repeating the image selection in the thumbnail unit 601 and the teaching process in the teaching area 602. In addition, the image data taught is stored in the evaluation image storage part 105.

N number of defects registered in the image data for evaluation and M number of imaging condition sets for one imaging condition set selected by setting respective setting values for each parameter item (1) to (5). In this case, the number of defects to be taught is N × M, and when M is large, registering all of them from the screen becomes impractical. As a method corresponding to this, the teaching of all of the image data acquired by the image capturing unit 101 is not performed, but the teaching data registered with respect to the image data acquired under any of the imaging conditions is acquired under other imaging conditions. It is also possible to apply it to image data.

As a specific procedure, first, ID is attached | subjected to each defect on a sample beforehand. Then, N defect image sets are obtained from the sample wafer by one of the imaging condition sets. Next, the defect positions are registered for these N images using the method shown in FIG. Next, the image of the same defect is imaged in other imaging conditions. At this time, only a defect having the same ID as the previously picked-up defect is picked up. By using this ID, it becomes easy to select the same defect from the defect image data acquired by different imaging conditions. Next, the defect image which has not yet been taught is selected, and the ID of the defect is acquired. Then, the completed defective image corresponding to the ID is obtained.

It is natural that the images of two identical defects acquired under different imaging conditions differ in image quality due to differences in imaging conditions, but in addition, there is usually a subtle visual field deviation caused by an error in stage stop accuracy or the like. Therefore, by detecting the pattern deviation of the visual field between the image that has been taught and the image of the same defect that has not yet been taught by pattern matching, and adding the shift amount to the defect position of the taught image that has not been taught, To estimate the location of the defect. Estimation of the defect position by this is performed for M set of picked-up images. As a result, only the teaching process is performed N times, and as a result, the defect position of the defect image data of the N × M sheets can be obtained.

Next, one condition is selected from the set of imaging conditions (S3), and defect detection processing is performed on (N pieces) the image data sets picked up under the conditions (S4). This defect detection process is performed by the defect detection execution unit 108 in the recipe evaluation unit 106. The defect detection execution unit 108 stores a program for executing defect detection processing therein, and is similar to the defect detection processing T5 executed in the processing flow of FIG. 3 with respect to the image of the wide field of view to be input. Has the ability to run the process offline. This defect detection process is performed for all N pieces of image data sets.

Next, the defect detection performance calculation part 109 of the inside of the recipe evaluation part 106 in the defect observation apparatus shown in FIG. 1 currently calculates the defect detection performance using the result data of these N defect detection processes. Throughput calculation unit 110 calculates throughput performance under selected imaging conditions, respectively (S5). This processing is performed for all M sets of imaging conditions.

In the calculation of the defect detection performance in the defect detection performance calculation unit 109, the defect detection position of the processing result in the defect detection execution unit 108 and the taught defect detection position are compared, and the difference in the position is evaluated. do. As the evaluation method, for example, there is a method for determining that defect detection succeeds if there is a defect detection position inside the circle 604 of the definition area defined at the time of teaching, and failure if it does not exist. For all N defect image data sets under the same imaging condition, it is determined whether success or failure occurs, and the ratio of the success number is regarded as defect detection performance.

On the other hand, in the calculation of throughput, the function data is used for the processing time stored in the processing time data storage 107. 7 is an example of such processing time data. In FIG. 7, the processing time at the time of providing each candidate value of the imaging condition parameter shown in FIG. 5 about each process step T1-T6 of the defect image collection shown in FIG. 3 is represented in table form. With respect to the currently selected imaging condition, the time of each processing from T1 to T6 is read from the table shown in FIG. 7 from the value of the setting parameter, and one defect is obtained by calculating the sum of the six values. The time for collecting the image of is calculated.

For example, the imaging conditions currently selected are as follows.

(1) Image size of the wide-field reference image: 1024

The number of additions of wide field reference image: 8

(2) Image size of narrow field defect image: 512

ㆍ Additional number of sheets of narrow field defect image:

In this case, T1: 600 msec, T2: 800 msec, T3: 400 msec, T4: 800 msec, T5: 1000 msec, and T6: 400 msec, so the total processing time is 4000 msec. By converting this value, the throughput (for example, the number of defects that can be automatically observed in one hour) is calculated to be about 900.

Since the data of each processing time shown in FIG. 7 is unique to the apparatus, it can be determined in advance. In this case, what is necessary is just to set an average value as a value. For example, the time required for the stage movement is the time required for the stage movement between the position of the defect portion for a certain defect and the reference portion of the other defect, and is strictly related to the distance between the defects, that is, the defect distribution. Therefore, although it changes, it is assumed that general reference | standard (for example, average moving distance = 10 mm etc.) is set and average time is shown.

In addition, although the method of estimating by accumulating based on each processing time data as shown in FIG. 7 as a calculation method of a throughput was shown here, in addition, for example, in process step S1, At the time of image data collection for evaluation, the time required for image collection may be measured and the value may be used.

Finally, the defect detection performance and the throughput of the various sets of imaging conditions calculated by the recipe evaluator 106 are output to the input / output unit 103. 8 is an example of the display screen. For each set of imaging conditions, the values of the parameters to be set, defect detection performance, and throughput performance are displayed in alignment. On this screen, data can be rearranged using any parameter as a key, and comparison between condition sets can be easily performed. By viewing these display results, selecting a plurality of conditions actually used (displayed with a check box), and clicking the registration button, the selected condition setting contents are stored in the recipe storage unit 104.

In addition, the condition setting registered in the recipe storage part 104 is not only instruct | indicated by a manual on the display screen in the input / output part 103 as mentioned above, but is based on the criteria which the recipe evaluation part 106 itself predetermine. It is also possible to determine automatically based on this. For example, "the defect detection performance is an imaging condition of 95% or more, and the throughput performance is the highest." In this case, on the display screen as shown in Fig. 9, only the condition that satisfies a predetermined criterion for the defect detection performance is highlighted in advance (in the example of Fig. 9, all the columns of the corresponding items are filled in gray and displayed). When the conditions with the highest throughput among those highlighted are automatically displayed in a state where a check mark is automatically displayed, confirmation by the operator becomes easier. In this case, it is also possible for an operator to register with a recipe storage part after confirmation, and to make it automatically register, without performing a confirmation operation. In any case, as shown in Fig. 8, for a plurality of imaging conditions, two performance indices of the defect detection performance and the throughput performance are displayed, and the function of selecting and sorting based on the condition parameter and the performance indices is further described. By selection, the selection of imaging conditions suitable for automatic collection is facilitated.

<Example 2>

Next, a second embodiment of the defect observation method according to the present invention will be described.

In the first embodiment, collection of evaluation data used for evaluation of defect detection performance has been performed through image capturing in the image capturing unit 101 according to a plurality of sets of imaging conditions set in advance. In the collection of evaluation image data, a review SEM according to the present invention having a function of simulating and evaluating evaluation data under different imaging conditions using image data that has already been collected, rather than actually capturing all data in this manner. Will be described in the present embodiment.

The apparatus structure of the review SEM of a present Example is shown in FIG. In the apparatus configuration shown in FIG. 1, a reference image storage unit 1001 that stores a reference image, which is the basis of image creation, in association with its imaging conditions, and an image generation unit 1002 that simulates and generates an image from the reference image. ) Is further provided. In the image generating unit 1002, whether or not the imaging condition generating unit 1003 for generating other imaging conditions and images of those conditions need to be imaged by the image pickup unit or simulation based on the imaging conditions of the reference image. It is comprised by the determination part 1004 which determines whether it is possible to create, and the simulator 1005 which performs image generation by simulation.

11 shows a processing flow according to the present invention. This example differs from the processing flow in the first embodiment shown in FIG. 4 with respect to a method of collecting an image set under different conditions and a teaching flow of a defective area.

First, list data showing information on whether or not image capturing is necessary in the image capturing unit in order to capture a list of the set value candidates for each parameter set as the imaging condition in advance, and an image in which the set value is changed for those parameters. Is stored in the reference image storage unit 1001. 12 is an example. The types of parameters are five types of acceleration voltage, probe current, average number, image size, and field of view size, and it is assumed that there are three kinds, four kinds, four kinds, four kinds, and four kinds of setting candidate values.

In addition, in FIG. 12, the image acquisition is shown as either "necessary" or "unnecessary" for each parameter. This indicates whether or not the image acquisition can be performed again when acquiring the image which changed the imaging condition from the actually imaged image. For example, it is "not required" about the parameter of average number of sheets. When the average number of sheets is changed, the S / N of the image changes, but since the change of the S / N can be simulated, when an image obtained by changing only the average number of conditions is obtained for an image acquired under a certain imaging condition, the actual image is obtained. This means that the simulation can be utilized without imaging.

And based on the parameter candidate list shown in FIG. 12, the reference imaging condition which becomes the basis of imaging condition generation is determined. At this time, the parameter for which image acquisition is necessary or not, specifically, the acceleration voltage, the probe current, and the field of view size, is set to a preset value. On the other hand, with respect to a parameter for which image acquisition is required or not, specifically, the average number and the visual field size, the average number of sheets is as many as possible (for example, when there is a parameter set value candidate list as shown in FIG. 12). In addition, as many as 16 of an upper limit and the image size are made as many as possible (in the case of FIG. 12, an upper limit of 1448 pixels). This is because, once the image data at such an upper limit value is acquired, it is easy to simulate and generate the imaging data under other conditions.

The flow of processing will be described in accordance with the flowchart shown in FIG. First, the evaluation image data is acquired by the image pickup unit 101 under reference imaging conditions (S1101). Next, the acquired position is taught through the display screen shown in FIG. 6 similarly to the first embodiment (S1102). Thereafter, a plurality of imaging conditions are set by combining the candidate values of each parameter shown in FIG. 12 (S1103). Next, one set image pick-up condition is selected (S1104), and image pick-up is performed or the image is generated in a simulation.

Here, determination (S1105) of imaging or simulating an image is performed as follows.

Step 1: The types of parameters with different setting values are acquired under the currently selected imaging conditions and reference imaging conditions. This parameter type is not limited to 1, but may be plural.

For the parameter extracted in Step2: Step1, the contents of "necessity of image acquisition" are acquired for the parameter from the table shown in FIG.

As a result of Step3: Step2, if none of the parameters are "needed" for "image acquisition necessity", simulation is created; otherwise, image pickup is performed.

In the case shown in Fig. 12, for example, when the acceleration voltage is included in the imaging condition parameter extracted in Step 1, acquisition by the image capturing unit 101 is required, and the extracted parameter is one, In the case of either the average number of sheets and the image size, and two parameters, and the average number of sheets and the image size, creation by simulation is performed.

The method of simulation generation process (S1106) is shown below. First, the creation of a simulation image relating to the average number of copies is as follows. Decreasing the average number means that the noise N is increased when the signal S is made constant in which the S / N ratio is lowered. Therefore, by performing a process of applying random noise to the obtained reference image, it is possible to simulate and create an image when the average number of sheets is smaller.

There is a statistical relationship between the average number of sheets and S / N that "when the average number is doubled, S / N increases by 2 times." Therefore, using this relationship, the simulator 1005 is obtained so that the S / N of the expected simulation image is obtained from the value of S / N calculated from the reference image and the value of the designated average number of sheets, and such an image is actually obtained. Is achieved by adjusting the amount of noise to be applied. In addition, the simulation regarding the image size generates an image of, for example, 1024 and 724 pixels by thinning the reference image (image size 1424). In the thinning process, it is important that the image S / N does not change before and after thinning.

On the other hand, when it is determined that imaging by the image capturing unit 101 is necessary for image acquisition of the selected imaging condition, image capturing is actually performed under the condition (S1107). In the image re-photographed in this way, since a defective image in which the teaching data has already been set is likely to be captured due to a shift in the field of view, similarly to the method shown in Example 1, the amount of field shift is detected by pattern matching. By adding the value, the defect position on the acquired image is identified and made into teaching data.

In addition, when a simulation is made, since there is no visual field shift, the teaching result with respect to a reference image can be used as the teaching result of a simulation preparation result as it is. In the process flow of FIG. 11, processes S1109 to S1112 after S1106 and S1108 are the same as processes S4 to S7 according to the first embodiment described in FIG. 4.

In this way, the image imaging unit 101 does not capture all the image data sets under different imaging conditions, and substitutes the simulation data according to the type of the parameter to be changed, thereby improving the efficiency of the image set creation. Since there are many kinds of recipes to be produced in a line of small quantity production of many kinds, the effect of the efficiency of image collection time related to the production of such one recipe is large. In the most prominent case, when acquiring an image whose parameters have been changed only by the conditions under which simulation can be created, i.e., only the number of copies and the image size, image capturing is completely unnecessary. Off-line operation, for example, at a terminal separate from the network, is possible.

<Example 3>

Next, a third embodiment of the defect observation method according to the present invention will be described. In the first and second embodiments described so far, only the defect positions were taught in the teaching process for the defective image data set for evaluation. This embodiment differs from the previous embodiment in that defect information other than the defect position is set.

On the teaching process screen (Fig. 6) described above, the defect information input unit 606 is an area for inputting various defect information. For each defect, at the time of registration of the defect position in the teaching area 602, information on the defect, its kind (with foreign materials, wiring short circuit, etc.), size, uneven state of the surface, brightness of the image, shape, etc. It is possible to register by inputting any one or more of the features.

In this way, when defect information other than the defect position is given to each defect, it becomes possible to set imaging conditions suitable for the automatic collection of defects according to their attributes. For example, it is possible to classify the evaluation sample into the defect type and the defect size, and to evaluate the relationship between imaging conditions, defect detection performance, and throughput for each classification result. For example, taking the defect size as an example, it is possible to divide the defect into two classes based on arbitrary sizes (for example, 1 micrometer), and to set appropriate imaging conditions for each class. For example, when a defect size is large, defect detection by image processing is generally easy, so for a class with a defect size larger than a reference value, even in a high throughput condition such as a small number of average copies or a small image size The risk of lowering the defect detection rate is low. On the other hand, defect detection becomes more difficult as it becomes a smaller defect, and therefore, for a class whose defect size is smaller than the reference value, it is preferable to perform image collection under conditions where the performance of defect detection is less likely to decrease, that is, at the expense of throughput. do.

13 shows a display screen of a result of evaluating defect detection performance and throughput for each class of defects. In this figure, the imaging condition evaluation result for the defect of size 1 micrometer or more is shown at the upper end, and the imaging condition evaluation result for the defect of size less than 1 micrometer is shown at the lower end. In each result, the same number given under imaging conditions is the same.

Since the ease of defect detection differs depending on the size of the defect size, when the defect size is large, 95% of defect detection performance is achieved under three conditions (conditions 2, 3, and 4) among the four conditions. In the case where the maximum throughput is 2250 and the defect size is small, only the condition 4 can achieve 95% of defect detection performance, and the throughput is lowered to 780.

Since the defect inspection apparatus has a function of outputting an approximate value of the size in addition to the position of the defect, by using this result, it is possible to set imaging conditions for automatic image collection suitable for the defect size for each defect. Some defect inspection apparatuses have a function of outputting the result of automatic classification of defects in addition to the defect size. Using such a result, switching of imaging conditions using not only the defect size but also the information of the defect type is possible.

101: image pickup unit
102: whole control unit
103: input and output unit
104: recipe storage
105: image storage for evaluation
106: recipe evaluation unit
107: processing time data storage
108: defect detection execution unit
109: defect detection performance calculation unit
110: throughput calculation unit
201: Sample
202: stage
203: electron source
204: electron beam
205: Condenser Lens
206: objective lens
207: Detector
208: Analog to Digital Converters
209 Deflector
210:
601: thumbnail part
602: teaching area
603: defect site
604: Definition area
606: defect information input unit
1001: reference image storage
1002: image generating unit
1003: imaging condition generation unit
1004: determination unit
1005: simulator

Claims (10)

Image pickup means for picking up a sample and obtaining an image of the sample, image pickup condition storage means for storing image pickup conditions for picking up the sample by the image acquisition means, and image pickup conditions stored in the image pickup condition storage means. A defect observing apparatus provided with a control means which controls the said image acquisition means based on and acquires the image of the desired part of the said sample,
The control means controls the image acquisition means to capture a plurality of images acquired by imaging a defect of a designated portion on the specimen under a plurality of imaging conditions, the plurality of imaging conditions, and defect detection performance for each of the plurality of imaging conditions. Image data storage means for storing;
Processing time data storage means for controlling the image acquisition means by the control means to associate processing time data required for imaging required for each of the plurality of imaging conditions when imaging the defect on the sample under a plurality of imaging conditions; and,
At the time of observing a defect on the sample using the imaging time conditions of the plurality of images stored in the image data storage means and the processing time data required for imaging corresponding to each of the imaging conditions of the plurality of images stored in the processing time data storage means. Performance evaluation means for obtaining throughput performance and defect detection performance of
Display means for displaying the throughput performance and the defect detection performance obtained by the performance evaluation means in association with the imaging conditions.
Defect observation device having a. The method of claim 1,
The image acquisition means is a scanning electron microscope, and the plurality of imaging conditions include any one of an image size, an image field size, a frame addition number of detected images, an acceleration voltage of an electron beam, and an amount of beam current. Fault Observation Device. The method of claim 1,
The processing time data includes a time for driving a stage on which a sample is mounted and a time for capturing the sample by capturing the sample. The method of claim 1,
The image data storage means is provided with information on a defect type of the defect in a plurality of images of the defect on the sample acquired by the image acquisition means, and stored by the performance evaluation means. A defect observing apparatus characterized by calculating defect detection performance and throughput performance. The method of claim 1,
The control means controls the image acquisition means to image a defect of a designated portion on the specimen and acquires an image of the defect. The area including the portion containing the specified defect on the specimen is included in the image acquisition means. Imaging at a first viewing size and acquiring an image of the first viewing size, detecting a defect position from the acquired first viewing size image, and acquiring the image smaller than the first viewing size with respect to the detected defective position Obtained by imaging an image with the 2nd visual field size of a means, The defect observation apparatus characterized by the above-mentioned. As a method of observing a defect on a sample,
The defect detection rate for each said imaging condition is computed from the image data obtained by imaging the defect on the said sample by changing the imaging conditions of the imaging means, and the defect position data which stored the position information of the defect in the image data,
Throughput for each imaging condition is calculated from the processing time data required for image acquisition corresponding to each imaging condition of the imaging means,
The calculated defect detection performance and throughput performance for each imaging condition are displayed on the screen,
And a defect on the sample is observed using the imaging means based on the imaging conditions selected on the screen on which the defect detection performance and the throughput performance are displayed. The method according to claim 6,
A scanning electron microscope is used as the imaging means, and any of the image size, the image field size, the number of frame additions of the detected image, the acceleration voltage of the electron beam, and the amount of beam current are taken as imaging conditions by the scanning electron microscope. Defect observation method comprising a. The method according to claim 6,
The processing time data includes a time for driving a stage on which a sample is mounted and a time for capturing the sample by capturing the sample. The method according to claim 6,
The image data includes information on a defect type of the defect in a plurality of images of the defect on the specimen, and calculating throughput for each imaging condition calculates defect detection performance and throughput performance for each defect type. Defect observation method characterized in that. The method according to claim 6,
The imaging conditions of the imaging means are changed to image the defects on the sample to obtain image data. The first field of view of the imaging means captures an area including a portion containing a specified defect on the sample at the first field of view. Acquiring an image of a size, detecting a defect position from the acquired first visual field size, and imaging an image at a second visual field size of the imaging means smaller than the first visual field size with respect to the detected defect position; Defect observation method comprising a.

Images